Introduction

I recently had a conversation with PJ Eby in the comments of my article “Building up to an Internal Family Systems model”. My most recent reply to it started getting rather long, and is also more broadly relevant as it contains updates on how I view IFS and the multi-agent framework in general. As a result, I decided to post it as its own article.

pjeby’s comments that I’m replying to are here and here; to verify that I understood him correctly, I wrote a summary of what I took to be his core points (below). He mostly endorsed them as correct, while having a few clarifications; the below summary has incorporated those corrections. After listing all of his points, I will present my own replies.

The summary is divided into two parts. Earlier on, pjeby wondered why IFS was popular among rationalists; one of the things I said in response was that rationalists like reductionism, and IFS helps reduce the mind into smaller components. pjeby felt that IFS is not good reductionism. My response goes into detail about what kinds of claims I view IFS as making, how I interpret those in terms of my multi-agent minds series, and how I would now rephrase my original article differently. Here I feel like I broadly agree with pjeby, and feel that our disagreement has more to do with using terminology differently than it does with actual object-level issues.

The second part concerns the practical usefulness of IFS as a therapeutic model. After all, a model can be useful while still not being true. Here I have more disagreement, and feel that (regardless of how good it is as a literal description of the mind) IFS has considerable practical value.

This article assumes that one has read my later article about Unlocking the Emotional Brain, as I reference concepts from it.

My summary of pjeby’s positions

IFS as a reductionist model

  • Good reductionism involves breaking down complex things into simpler parts. In the case of doing reductionism on agents, the end result should be mechanisms without agency. IFS "breaks down" behavior into mini-people inside our heads, each mini-person still being just as agenty as a full person is. This isn't simplifying anything.
  • Talking about subagents/parts or using intentional language causes people to assign things properties that they actually don't have. If you say that a thermostat "wants" the temperature to be something in particular, or that a part "wants" to keep you safe, then you will predict its behavior to be more flexible and strategic than it really is.
  • The mechanisms behind emotional issues aren't doing anything agentic, such as strategically planning ahead for the purpose of achieving a goal. Rather they are relatively simple rules which are used to trigger built-in subsystems that have evolved to run particular action patterns (punishing, protesting, idealistic virtue signalling, etc.). The various rules in question are built up / selected for using different reinforcement learning mechanisms, and define when the subsystems should be activated (in response to what kind of a cue) and how (e.g. who should be the target of the punishing).
  • Reinforcement learning does not need to have global coherence. Seemingly contradictory behaviors can be explained by e.g. a particular action being externally reinforced or becoming self-reinforcing, all the while it causes globally negative consequences despite being locally positive.
  • On the other hand, IFS seems to imply that there is dedicated hardware for each instance of an action pattern: each part corresponds to something like an evolved module in the brain, and each instance of a negative behavior/emotion corresponds to a separate part. Even if IFS does not assume it, then at least this article is assuming dedicated agents actually existing.
  • The assumption of dedicated hardware for each instance of an action pattern is multiplying entities beyond necessity. The kinds of reinforcement learning systems that have been described can generate the same kinds of behaviors with much less dedicated hardware. You just need the learning systems, which then learn rules for when and how to trigger a much smaller number of dedicated subsystems.
  • The assumption of dedicated hardware for each instance of an action pattern also contradicts the reconsolidation model, because if each part was a piece of built-in hardware, then you couldn't just entirely change its behavior through changing earlier learning.
  • Everything in IFS could be described more simply in terms of if-then rules, reinforcement learning etc.; if you do this, you don't need the metaphor of "parts", and you also have a more correct model which does actual reduction to simpler components.

IFS as a therapeutic approach:

  • Approaching things from an IFS framework can be useful when working with clients with severe trauma, or other cases when the client is not ready/willing to directly deal with some of their material. However, outside that context (and even within it), IFS has a number of issues which make it much less effective than a non-parts-based approach.
  • Thinking about experiences like "being in distress" or "inner criticism" as parts that can be changed suggests that one could somehow completely eliminate those. But while triggers to pre-existing brain systems can be eliminated or changed, those brain systems themselves cannot. This means that it's useless to try to get rid of such experiences entirely. One should rather focus on the memories which shape the rules that activate such systems.
  • Knowing this also makes it easier to unblend, because you understand that what is activated is a more general subsystem, rather than a very specific part.
  • If you experience your actions and behaviors being as caused by subagents with their own desires, you will feel less in control of your life and more at the mercy of your subagents. This is a nice crutch for people with denial issues who want to disclaim their own desires, but not a framework that would enable you to actually have more control over your life.
  • "Negotiating with parts" buys into the above denial and has you do playacting inside your head without really getting into the memories which created the schemas in the first place. If you knew about reconsolidation, you could just target the memories directly, and bypass the extra hassle.
  • "Developing self-leadership" involves practicing a desired behavior to override an old one; this is what Unlocking the Emotional Brain calls a counteractive strategy, and is fragile in all the ways that UtEB describes. It would be much more effective to just use a reconsolidation-based approach.
  • IFS makes it hard to surface the assumptions behind behavior, because one is stuck in the frame of negotiating with mini-people inside one's head, rather than looking at the underlying memories and assumptions. Possibly an experienced IFS therapist can help look for those assumptions, but then one might as well use a non-parts-based framework.
  • Even when the therapist does know what to look for, the fact that IFS does not have a direct model of evidence and counterevidence makes it hard to find the interventions that will actually trigger reconsolidation. Rather one just acts out various behaviors which may trigger reconsolidation if they happen to hit the right pattern.
  • Besides the issue with luck, IFS does not really have the concept of a schema which keeps interpreting behaviors in light of its existing model, and thus filtering out all the counter-evidence the playacting might otherwise have contained. A skilled IFS therapist paying attention might notice this happening and deal with it, but such a good therapist could probably get results with any therapy model that gave them sufficient freedom to notice and address the issue.

My responses

IFS as a reductionist model:

  • Good reductionism involves breaking down complex things into simpler parts. In the case of doing reductionism on agents, the end result should be mechanisms without agency. IFS "breaks down" behavior into mini-people inside our heads, each mini-person still being just as agenty as a full person is. This isn't simplifying anything.

I agree that IFS feels pretty anthropomorphic, and especially if you take the subpersonalities thing too literally, it’s going to guide you away from more correct models. And I agree that reducing things to subpersonalities isn’t really reducing things to simpler mechanisms.

That said, there are degrees of correctness and wrongness. We got into the discussion about reductionism when pjeby wondered why IFS was so popular among rationalists, and I suggested that it might be in part because rationalists like reductionism. Now, maybe reductionism was a bad term, and I should have used something like “rationalists like having better models" instead.

In any case, what I was trying to say was that even if IFS isn’t great reductionism, it’s still a better model than a naïve conception of the mind as a unified whole. Even if some aspects of it are misleading, “this person is doing self-destructive things because of a firefighter response to a triggered exile” is better than “this person is doing self-destructive things because they are too stupid or irresponsible to act better”.

I think the main virtue of IFS is just that it breaks down the mind into, well, parts, each containing different perspectives and goals. There are many interpretations of IFS; some go with the full-blown “subpersonalities” approach, whereas some just consider the subpersonalities a metaphor for something fundamentally non-agenty. Even the first edition of the Internal Family Systems Therapy textbook noted that while in the author’s experience, the system works better if you treat parts as independent entities, many IFS therapists just treat them as metaphors and that this works too:

To use the IFS model, one does not have to believe in an ontological sense that parts are internal people. Many therapists view this depiction merely as a useful metaphor and have success with the approach. From a pragmatic perspective, however, the inhabitants of our internal systems respond best to that kind of respect, so it is best to treat them that way—to attribute to them human qualities and responses. Some therapists resolve this problem by not viewing parts as people when they are theorizing, but seeing them as inner people when they are treating clients.

In our previous discussion, there were moments when I said something like “if you treat the mind as a reinforcement learning system, then that predicts X which doesn’t actually match human behavior, but if you treat the mind as an IFS-like system, then that makes more sense”. pjeby responded by pointing out that reinforcement learning does not need to be globally coherent, and things can actually be explained fine in a reinforcement learning framework if one postulates various local RL mechanisms. I think we were talking past each other, because an RL framework with local mechanisms is the kind of thing that I meant by "an IFS-like system".

The distinction that I was drawing was between “a unified top-down system of the kind that one finds in folk psychology or most contemporary AI designs” on the one hand, versus “a system made up of many distinct learning mechanisms and knowledge stores” on the other. When I said that IFS might be popular among rationalists because they like reductionism, I just meant that IFS is a good model for many because it naturally guides people towards thinking along the latter kinds of lines.

E.g. the most top-voted comment on my IFS article says that what it suggests “about procrastination and addiction alone [...] are already huge”, so even in 2019, many rationalists found the IFS model to be better at offering explanations than any other model. Here's another recent comment that seems to have moved towards a latter kind of model through the IFS frame:

Over this past year I've been thinking more in terms of "Much of my behavior exists because it was made as a mechanism to meet a need at some point."
Ideas that flow out of this frame seem to be things like Internal Family Systems, and "if I want to change behavior, I have to actually make sure that need is getting met."

Of course, pjeby’s point was that there are even better models than what one gets from IFS; and I agree. The IFS model was one of the starting points of my sequence, but my later ones have been moving more explicitly towards a UtEB kind of model. The intention of my comment was descriptive rather than prescriptive; that historically, the IFS model has been popular because it’s pragmatically useful and because, despite its possible flaws, rationalists haven’t been exposed to any better models.

  • Talking about subagents/parts or using intentional language causes people to assign things properties that they actually don't have. If you say that a thermostat "wants" the temperature to be something in particular, or that a part "wants" to keep you safe, then you will predict its behavior to be more flexible and strategic than it really is.

I agree that excessive anthropomorphization is a possible issue with both IFS in particular and intentional language in general. As pjeby says, it can cause people to misapply mental models derived from full-blown humans into parts of the psyche, preventing people from realizing what’s going on. (In fact, this was one of the main reasons why I stayed away from IFS for a long time, since it sounded like naïve folk psychology at best.)

Though I don’t think it’s necessarily a problem. When pjeby said

… if you actually want to predict how a thermostat behaves, using the brain's built-in model of "thing with intentional stance", you're making your model worse. If you model the thermostat as, "thing that 'wants' the house a certain temperature", then you'll be confused when somebody sticks an ice cube or teapot underneath it, or when the temperature sensor breaks.

… then it’s worth noting that the reason why this example works is that neither of us is actually going to make that kind of a mistake. We know enough about mechanical stuff to realize that what’s meant by “wanting” in this context, and what its limitations are.

Now again, it’s true that if you don’t know anything about mechanical stuff, then you will be misled. On the other hand, if you do know enough to understand what’s meant, then it’s a convenient shorthard and aid for thinking. Evolutionary biologists might talk about e.g. “evolution selecting for a trait”, which makes evolution sound like an agent which is doing things, even though they fully well know what’s actually happening.

In one comment, I said that the parts in IFS feel basically like the schemas in UtEB to me, and pjeby replied:

A "model" is passive: it merely outputs predictions or evaluations, which are then acted on by other parts of the brain. It doesn't have any goals, it just blindly maps situations to "things that might be good to do or avoid". An "agent" is implicitly active and goal-seeking, whereas a model is not.

But suppose that we have these descriptions:

Agenty: A manager part scans the environment for things it considers threats, taking a particular action whenever it detects something that it considers a potential sign of threat.

Passive: A model contains a record of some kinds of sensory inputs which have previously been associated with negative consequences, and also contains an action to be taken in response. When cues matching that record are detected in the system’s sensory input, the system takes the corresponding action.

… then the “passive” version sounds to me like it’s just a description of how the “agenty” version is implemented. Especially if one knows (as I assume LW readers to know) that there’s no clear-cut distinction between code and data, or that the distinction is even messier in the brain.

The most famous society of mind model is probably Minsky's Society of Mind; and while it seems to define "agents" in different ways in different chapters, some of them seem very low-level. I recall one chapter which sounded a lot like the "agents" in it were individual nodes in a neural network; another chapter had agents like "the add operation". So it doesn't feel like talking about "agents" would necessarily imply a lot of agency.

But then on the third hand, I again agree that the potential for being misled is still there - especially since I got slightly misled myself, when I wrote my IFS post and formulated things in terms of the “subagents” themselves doing reinforcement learning. I’ve been moving away from that kind of a formulation in my later posts, and would write this post differently if I were to do it now. Maybe I should change the name of the sequence from “multiagent minds” while I’m at it, but then I would need to think of some equally catchy name. :-)

On the fourth hand, there are things in my IFS post which, despite being framed in an agenty way, I think are still formally equivalent to a more passive formulation. E.g. I talked about subagents voting for different objects they wish to become the content of consciousness, and the winner being randomly determined based on the vote counts; this still seems like a fair way of describing things like winner-take-all networks or biased competition, assuming stochasticity in the process.

  • The real mechanisms behind emotional issues aren't really doing anything agentic, such as strategically planning ahead for the purpose of achieving a goal. Rather they are relatively simple rules used to trigger built-in subsystems that have evolved to run particular kinds of action patterns (punishing, protesting, idealistic virtue signalling, etc.). The various rules in question are built up / selected for using different reinforcement learning mechanisms, and define when the subsystems should be activated (in response to what kind of a cue) and how (e.g. who should be the target of the punishing).

I like this model. It’s similar to what I’ve been moving towards in my more recent posts, and reminds me of e.g. Janina Fisher’s model, but is more precise about what exactly is happening.

  • Reinforcement learning does not need global coherence. Seemingly contradictory behaviors can be explained by e.g. a particular action being externally reinforced or becoming self-reinforcing, all the while it causes globally negative consequences despite being locally positive.

Yes, I agree. In e.g. my post on subagents and coherence, I specifically talked about the overall system doing learning on which subsystems to activate in different situations, based on local considerations such as which subsystems have previously made accurate predictions. And my post on neural Turing machines tried to sketch out some of those mechanisms.

When I said that reinforcement learning alone couldn’t explain some behaviors, I was again drawing the distinction between “a unified top-down system” vs. “a system made up of many distinct learning mechanisms and knowledge stores”. I was interpreting “reinforcement learning” as an instance of the former, because my model of the mechanisms behind IFS already involves the kinds of RL mechanisms that pjeby has been describing.

  • On the other hand, IFS seems to imply that there is dedicated hardware for each instance of an action pattern: each part corresponds to something like an evolved module in the brain, and each instance of a negative behavior/emotion corresponds to a separate part. Even if IFS does not assume it, then at least this article is assuming dedicated agents actually existing (whether in software or hardware).
  • The assumption of dedicated hardware for each instance of an action pattern is multiplying entities beyond necessity. The kinds of reinforcement learning systems that have been described can generate the same kinds of behaviors with much less dedicated hardware. You just need the learning systems, which then learn rules for when and how to trigger a much smaller number of dedicated subsystems.
  • The assumption of dedicated hardware for each instance of an action pattern also contradicts the reconsolidation model, because if each part was a piece of built-in hardware, then you couldn't just entirely change its behavior through changing earlier learning.
  • Everything in IFS could be described more simply in terms of if-then rules, reinforcement learning etc.; if you do this, you don't need the metaphor of "parts", and you also have a more correct model which does actual reduction to simpler components.

I agree that “dedicated innate hardware for each instance of an action pattern” seems wrong, and that the same thing is more easily explained by learned rules. Again, if I were to re-write my IFS article now, I would model things somewhat differently. Though again, I don’t see there being that big of a difference between “a dedicated agent” and “a dedicated set of rules”, once you have a correct understanding of what “agent” means. And when you do have that understanding, you might as well use the “agent” term as a convenient shorthand.

Honestly, IFS doesn’t seem to be really clear on exactly how specific/general the parts are interpreted to be. To some extent, this might be because of IFS’s general philosophy of investigating each client as a unique system and not imposing too much of a pre-existing theoretical framework on them.

The second edition of the Internal Family Systems Therapy book has a mention of an infant researcher who “observed infants rotating among four or five discrete states, which we would call parts”, suggesting a stance of interpreting parts as pretty general brain systems; likewise when it suggests that a new part has come online when a previously compliant 2-year-old wakes up one day and insists on saying “no” to everything. But then in practical work, a lot of the parts seem a lot more specific.

My guess is that the “interface” used for creating the appearance of parts isn’t really that specific and may sometimes target narrower and sometimes broader wholes; e.g. it might access one specific schema or set of experiences, and sometimes a collection of interrelated schemas (referred to when IFS claims that “parts have parts”). Still, it seems still reasonable to me to think that IFS's parts generally correspond to something like UtEB's schemas.

Also worth noting that the second edition of Internal Family Systems Therapy explicitly references the memory reconsolidation model as a possible explanation of how parts are healed:

Our colleague Frank Anderson (2013) has pioneered using IFS in psychopharmacology. In the manual on IFS that we coauthored with Frank, he speculated that the IFS steps of witnessing and unburdening may correlate with a process neuroscientists call memory reconsolidation, a form of neuroplasticity described by Ecker, Ticic, Hulley, and Neimeyer (2012) that “changes existing emotional memory at the synaptic level” (Anderson et al., 2017, p. 127). In contrast to memory reconsolidation, a therapy like cognitive behavioral therapy (CBT) uses counteractive change strategies, which focus on developing new neural networks to counter older neural networks. As described in the manual, memory reconsolidation includes four phases. The first involves getting access to the traumatic memory. In IFS this would be finding, focusing, and fleshing out the target part. The second phase involves reactivation or destabilizing the emotional memory network and unlocking it at the synaptic level. In IFS this would be unblending, in which parts make room for the client’s Self. The third phase involves the mismatch, which is “a full disconfirmation of the meaning of the target memory” (Anderson et al., 2017, p. 127). In IFS this would occur during the process of witnessing and retrieving exiled parts. Finally, the fourth phase, erasure, involves clients making use of a new perspective to amend their understanding of the traumatic experience. In IFS this would happen when exiles let go of their burdens because the Self has a validating, nontraumatic perspective on the meaning of the part’s experience: You’re not bad—a bad thing happened to you.

The practical usefulness of IFS as a therapeutic approach

  • Approaching things from an IFS framework can be useful when working with clients with severe trauma, or other cases when the client is not ready/willing to directly deal with some of their material. However, outside that context (and even within it), IFS has a number of issues which make it much less effective than a non-parts-based approach.
  • Thinking about experiences like "being in distress" or "inner criticism" as parts that can be changed suggests that one could somehow completely eliminate those. But while triggers to pre-existing brain systems can be eliminated or changed, those brain systems themselves cannot. This means that it's useless to try to get rid of such experiences entirely. One should rather focus on the memories which shape the rules that activate such systems.

Note that IFS doesn’t really advocate getting rid of e.g. anger either, but rather transforming angry parts so that they are less extreme. That is, don’t get angry when it would be a bad idea, only get angry when it’s a good idea. This seems basically the same as what pjeby is saying: change the rules which regulate how the different systems are triggered.

Of course, it’s still possible to get the impression that one could somehow entirely eliminate particular response patterns. But is the other model any safer from a similar interpretation? If the IFS-user-who-never-wanted-to-be-angry thought “I will transform all of my angry parts so that I’m never angry anymore”, then the UtEB-mindhacker-who-never-wants-to-be-angry might think “I will eliminate all of the rules which trigger the angriness subsystem, and then that system will never get triggered anymore”.

I would expect both to fail for the same reason (being motivated by an anti-angriness schema rather than a desire to actually find the best response for any given situation), but I don’t see the UtEB-style framework offering any particular defense against this failure mode.

  • Knowing this also makes it easier to unblend, because you understand that what is activated is a more general subsystem, rather than a very specific part.

I tried this a bit the other day when I had some mild anxiety, intentionally reframing the experience as a subsystem being activated rather than a part. I think it might have made it slightly easier to unblend, but it was honestly probably mostly placebo effect from trying something new? Not sure why one would expect it to make a difference which frame you use; maybe if you thought of the part as actively fighting back or something, then that would introduce extra effort into the unblending process. But in both frames, you are choosing to interpret the reaction as something distinct from “you”, and paying attention to features which let you distinguish it from “yourself”. That seems to be crucial ingredient, rather than one’s interpretation of what exactly it is that one is unblending from.

  • If you experience your actions and behaviors being caused by subagents with their own desires, you will feel less in control of your life and more at the mercy of your subagents. This is a nice crutch for people with denial issues who want to disclaim their own desires, but not a framework that would enable you to actually have more control over your life.

I guess this could be a problem if you experienced your subagents as being really strong and independent and hard to reason with. But the IFS notion of all parts - even the most malicious and “evil-seeming” ones - fundamentally wanting your best, and its super-cooperative attitude towards them, seems to be good at framing things in a way that avoids this impression. If anything, people often feel liberated when they use IFS techniques and stop experiencing a part as something that torments them and who they are at the mercy of, and come to view it as a friend and an ally instead.

Of course, if one does have a schema that guides them into strong helplessness behaviors or into denial, then they can just filter out those parts of the IFS framing and get something like the denial-supporting model described. But again, I’m not sure if there’s anything in another framework either that would prevent one from forming an equivalent interpretation? Instead of feeling at the mercy of your subagents, you would feel at the mercy of your schemas and rules for triggering subsystems.

Introspecting on my own experience, I feel like the rule-based framing might actually put me more into the “denying my own desires” mode. It primes me to think along the lines of “oh, all of my bad behaviors are just caused by bad irrational rules, I’ll use consolidation techniques to rewrite all of them”. Whereas if I model myself as IFS-like parts, then I think more along the lines of “hmm, all of my parts have some good intention for me and some reason for acting the way they do, let me see if they might have some important insight that I’ve missed”. (Which, in my experience, is a stance that gets much better results. SaraHax reports similar things.)

  • "Negotiating with parts" buys into the above denial, and has you do playacting inside your head without really getting into the memories which created the schemas in the first place. If you knew about reconsolidation, you could just target the memories directly, and bypass all of the extra hassle.

This doesn’t really describe my experience of IFS. My feeling (both from doing IFS personally and facilitating it to others) is that it is not so much playacting, but rather basically just Focusing with a slightly different frame. Mark Lippmann describes IFS as “a structured super-charger for Focusing”, and notes that Focusing, IFS and Coherence Therapy seem to all be doing basically the same thing, with slight variations in the framing and questions that are used giving you radically different answers.

Let’s go with UtEB’s model of schemas containing a problem, solution, and memories that formed the schema. Again in my experience, “protectors” seem to correspond to the problem and solution parts of the schema, and “exiles” to the original experiences giving rise to that schema.

pjeby says that “if you knew about reconsolidation, you could just target the memories directly”, but AFAICT, the questions that IFS suggests you ask your protectors are intended to give you access to the memories directly. One is asking questions like “what is the goal of this part”, “what is the part afraid that would happen if it didn’t do its job” (note that this one is basically the same as what Richard was asked to imagine in the UtEB example), and “what is this part’s core belief”. These are all queries for eliciting more information about the schema in question; the reason why one fleshes out a part's appearance is to get a stronger felt sense handle to run those queries against.

“Negotiating with parts” seems to likewise be about passing information between conflicting schemas (this is more explicit in CFAR’s IDC technique, which is basically IFS without the stuff about healing exiles). One schema says that you should go into a particular situation, another says that it’s dangerous. So you elicit information from the schema which knows about the dangers and pass it to the pro-active schema, reconsolidating its model of “going to this situation would be a pure win”, and getting a more nuanced prediction about how the danger could be avoided. Then you pass it back to the objecting schema, reconsolidating it in turn, until the two schemas return a compatible prediction and you know how you should proceed.

IFS claims, and this has often also often been my experience, that if you can access the exile (the original memories behind the schema) from a place of Self, then healing will happen naturally and intuitively. pjeby suggests that this happens mostly accidentally, if the person happens to hit upon the right behaviors to trigger the reconsolidation. But my experience doesn’t support that: if I have proper access to an exile and I’m truly in Self, the right behaviors come intuitively and reliably (maybe with just a bit of nudging from the therapist/facilitator), as IFS suggests.

My model is that the mind is wired to avoid discomfort, and that schemas tend to contain the implicit assumption of “if memories relating to this experience are triggered, then that will always make me feel horrible, so I need to avoid that”. For example: you weren’t polite to your parents and you got shamed for it. This is a problem to be solved (in the Coherence Therapy sense): impoliteness will lead to social disapproval. The knowledge of this problem is neurally encoded by associating the feeling of shame together with the memory/concept of being impolite.

My key assumption here is that unlike the logical brain, the emotional brain does not have a concept of social disapproval being bad for instrumental reasons. Rather, social disapproval is bad because it triggers bad feelings, and the emotional brain is optimizing for the avoidance of bad feelings rather than for the avoidance of objectively observable circumstances. Thus, on the evolutionary level, brains are wired to avoid social disapproval because it is bad for survival; but on an algorithmic level, the emotional brain is wired to avoid social disapproval because social disapproval causes shame (presumably produced by some separate shame module).

So the concepts of impoliteness and shame become linked together to create a self-fulfilling prediction: experiencing yourself as impolite triggers shame through the link. The brain's reinforcement learning machinery comes to anticipate this and starts developing strategies for avoiding that experience. This prediction of (impoliteness -> shame) is reinforced each time when you experience or remember being impolite, and feel the corresponding emotion.

But when you experience the original painful memory from a place of Self, then you are recalling it without also feeling horrible. Similar to Richard noticing that confidence does not automatically lead to social rejection, this is counterevidence to the expectation of impoliteness always causing you to be shamed. This acts as a juxtaposition experience, eliminating the system's need to compulsively avoid impoliteness... and in so doing, reconsolidates the concept of impoliteness so as to remove the link to feelings of shame, meaning that the concept will no longer trigger shame.

E.g. in my IFS training, it was remarked that when parts say that the result of them not doing their job properly would be some concrete external consequence, one should keep asking for “and what would be so bad about that.'' In other words, keeping asking until you get a concrete prediction of an emotion or feeling as a result; that’s the thing that the schema is really trying to avoid, not the "objective" consequence. When you get that prediction, then you can activate it and provide it with counterevidence from the fact that you are accessing it from Self, and thus not feeling that horrible after all.

Now, I have had IFS sessions which seemed to involve “playacting” and seemed to go along a strong narrative arc, which I can’t really put in any sensible memory reconsolidation terms. It still seemed to do something, but didn’t feel super-effective. Probably another approach would have been more efficient there. So I am not claiming that IFS is always the best way to approach things: it’s probably effective for some things and less effective for others.

  • "Developing self-leadership" involves practicing a desired behavior so that it could override an old one; this is what Unlocking the Emotional Brain calls a counteractive strategy, and is fragile in all the ways that UtEB describes. It would be much more effective to just use a reconsolidation-based approach.

I wouldn't think of developing Self-leadership mostly as a counteractive strategy. It does have some counteractive components, such as learning to unblend when triggered. But that's mostly because any general strategy that you can use whenever you are triggered, before you have had a chance to use reconsolidative techniques or know enough about the schema in question to do it, has to necessarily be counteractive.

But the way I see it, the development of self-leadership is mainly reconsolidative as well. Take again UtEB's model where a schema contains both a problem description and a strategy for solving it. In my interpretation, “negotiating for Self-leadership” with a part is something you do when you have a grasp of the part’s (schema’s) strategy for solving a problem, but haven’t yet had the opportunity to dig into the underlying problem and associated memories. (Possibly because you only learned of this particular schema when you were triggered ten seconds ago, or because this is tricky in some other way.) In that situation, you apply reconsolidation to the strategy - basically showing the schema that even if it is responding to a valid problem, you can deal with it better if you are allowed to remain in Self.

E.g. here’s an excerpt from Internal Family Systems Therapy (1st ed.) describing a negotiation for Self-leadership; I’ve added comments in bold.

I asked Nina to imagine that she was at the graveyard with her mother-in-law and to describe her thoughts and actions. She said that her mother-in-law was telling her stories about Tom between fits of sobbing. Superwoman [a part] had taken control of Nina and was giving the tearful woman platitudes, with barely concealed impatience and discomfort. She felt distant from her mother-in-law and knew that she was not being very helpful, but believed she had to rely on Superwoman to keep her from falling apart in that context. [Up until now, the procedure has been very similar to the Coherence Therapy example where Richard was asked to imagine being in a meeting and describe what he expected to happen. There’s a situation which triggers a part/schema, and associated feelings and responses. Of course, CT didn’t use the “part” terminology, but that doesn’t make the essential content of the exercise different.]
I told Nina to talk with Superwoman and the other activated parts about letting her be her Self with her mother-in-law. They were to try not to interfere, and instead see how it went with Nina’s Self in the lead. The parts reluctantly agreed to watch instead of jumping in; Nina then imagined the scene at the cemetery again, but this time with the parts watching in the background instead of struggling to be in control. [Making your pre-existing implicit knowledge explicit by constructing a mental simulation which uses it; this technique featured in UtEB also.] Nina was amazed at how much more cooperative her parts were after she and they had done the internal work.
After a few minutes of silence, Nina reported that things had gone much better for both women. Nina and her mother-in-law had cried together, holding and stroking each other in commiseration. She felt closer than she ever had to the older woman, and was extremely relieved not to have to pretend to be so strong. [This makes for a juxtaposition experience: the prediction that Nina needs the Superwoman’s response to avoid falling apart, contrasted with the simulated result of actually being able to act better when in Self. As there is an update to the "Superwoman" schema's prediction that its strategy being activated will produce better results, it becomes less likely to trigger.] I asked Nina how her parts reacted to this; she said that they seemed impressed and surprised that it went so well, although Superwoman was still skeptical. Nina decided to try to repeat this experiment in vivo that week, and I wished her luck. [...]
At the [next] session, Nina proudly reported that her trip to the cemetery had unfolded almost as she had envisioned: She and her mother-in-law were able to cry together. She was also surprised at how nurturing her mother-in-law was able to be, once Nina gave her the chance. [after the strategy component of a schema has been successfully reconsolidated, the new behavior becomes effortless; if necessary, this can be used as additional evidence to further reconsolidate either the same schema or different ones]

IFS claims that as one does additional work, parts gradually “come to trust the Self” more, and it becomes easier to get parts to grant Self-leadership. My guess is that this is because you accumulate more and more evidence of it actually being a good idea to act out of Self, rather than from a triggered state. That will make it easier to find evidence to present to any schema which suggests acting otherwise.

Earlier pjeby said:

It's far more efficient to delete the rules that trigger conflicting behavior before you try to learn self-leadership, so that you aren't fighting your reinforced behaviors to do so.

And this is in fact also recommended in IFS; from the same discussion that I was quoting before:

Nina was amazed at how much more cooperative her parts were after she and they had done the internal work. [...] I find that this kind of in-sight (which I call the “Self-confidence technique”) is often useful for helping a client’s parts see that they can trust the Self. When parts let the Self lead in external situations, those situations are inevitably handled better than when a part leads. If parts refuse to allow the Self to lead, then their fears can be explored and addressed until they are willing to temporarily cede control to the Self.

In other words, if you can’t get Self-leadership to work, then dig into your schemas, see what they predict would go wrong, and do reconsolidation until it comes naturally.

  • IFS makes it hard to surface the assumptions behind behavior, because one is stuck in the frame of negotiating with mini-people inside one's head, rather than looking at the underlying memories and assumptions. Possibly an experienced IFS therapist can help look for those assumptions, but then one might as well use a non-parts-based framework.
  • Even when the therapist does know what to look for, the fact that IFS does not have a direct model of evidence and counterevidence makes it hard to find the interventions which will actually trigger reconsolidation. Rather one just acts out various behaviors which may trigger reconsolidation if they happen to hit the right pattern.

As I’ve noted, I feel like IFS is already all about looking at the underlying memories and assumptions. This is certainly what my IFS training emphasized: keep asking questions which will reveal the assumptions of the protectors, and take you to those exiles. The "UI" of the parts is used as an experiental / felt sense handle that leads back to the core of the schema.

It’s true that one might use a different framework as well, and sometimes that may have its benefits. But what makes IFS unique, I think, is that you don’t need to understand all the underlying assumptions of the schema: whatever they are, they bottom out at “and then I would feel terrible”, and if you can get to Self, you can just reconsolidate that assumption directly. And framing things in terms of parts seems to be useful in getting the person into Self, by asking parts to move aside.

  • Besides the issue with luck, IFS does not really have the concept of a schema which keeps interpreting behaviors in the light of its existing model, and thus filtering out all the counter-evidence that the playacting might otherwise have contained. A skilled IFS therapist who was paying attention might notice this happening and deal with it, but such a good therapist could probably get results with any therapy model that gave them sufficient freedom to notice and address the issue.

To finish this post off on an agreeable note: here I’m actually somewhat inclined to agree with pjeby. IFS as it’s taught does have something like this concept: it’s referred to as a “Self-like part”, i.e. a part which believes itself to be Self and tries to go through the motions of Self, but doesn’t actually get any real healing done. But even though IFS does have this concept, I’m not convinced it has very good tools for dealing with it. The main one that I heard of was just to insist that the part in question is not actually Self, and match the therapist’s conviction and reassurance against the part’s, until it moves aside and gives you access to the real Self. But I’m not sure of what one is supposed to do if that doesn’t work.

Thanks to Maija Haavisto for comments on drafts of this article.

New Comment
31 comments, sorted by Click to highlight new comments since: Today at 10:04 AM

Having addressed the reductionist side (where you basically agree with me, yet still think "agency" is useful), let's now look at the practicalities.

As far as I can tell, your practical argument in favor of IFS has three main parts:

  • Meta-schema (e.g. of seeking to exterminate bad parts/rules) are a thing that exists
  • IFS can arrive at the same place as other models by different steps
  • IFS' frame of positive intention is useful, and isn't available in other models

But none of these arguments actually favor IFS over other systems, or even distinguish it at all!

For the first argument, the thing I previously described as "The Interdict of Merlin for Self-Help" prevents people with meta-issues from resolving them without either 1) banging their head on a wall till they notice there's a wall there, or 2) having another living mind say, "hey, did you know that's a wall you're banging your head against?". In short, meta-issues are a Fully General Counterargument because they apply to any method of change for human brains. So they don't distinguish IFS from other systems.

Second, of course IFS can arrive at the same place as other models; my argument is that it lacks both rigor and simplicity in how it arrives at those places compared to models based on passive schema, that use an understanding of reconsolidation to explicitly focus on contradicting explicitly-selected memory targets.

And the authors of UtEB specifically note that therapies such as IFS work -- to the extent that they do -- because they are doing things that unintentionally or indirectly result in reconsolidation. So this second argument doesn't actually distinguish IFS from any other therapy that accidentally (i.e., without explicit intention/targeting) produces reconsolidation.

Finally, the third argument is, I think, where the real thrust of this entire article lies. I get the impression in fact, that this third argument is the reason you're defending IFS so strongly in the first place: that this frame has had personal meaning and utility to you.

But that "positive intention" frame isn't new, and definitely isn't unique to IFS. Other therapy modalities had it long before IFS was developed, and you can use a reductionist modality without losing the benefits.

For example, I might say to someone that something they learned to was the best thing they could do in the situation they were in, and back that up by helping them experience how and why that was the best thing, leading to the same sort of subjective experience as you describe, of realizing that your brain is not, in fact, out to get you.

And this framing doesn't require me to postulate a part that has your well-being in mind, yet for some reason keeps doing the same thing over and over! "You learned to do this because it was good then, and it's not so good now" is IMO a simpler theoretical frame than "you have parts that are well-intentioned, but also kind of dumb".

(Of course, the way I do this, there's an even more basic experience, of one's rules being "think X, feel Y automatically". By the time we're talking about anything like the positive intention frame, it's already been established by simple repetition that the troubling thing acts like a rule that wires a button directly to a feeling. So the question raised is not, "why is this part doing this to me?" but "how did I learn that rule?")

So the third argument doesn't distinguish IFS from other methods, and the subjective experience of unblending via positive intention can be arrived at by other means, that IMO are more efficient as well as more epistemically correct, and allow for greater rigor and speed.

IOW, on none of these three dimensions is IFS substantially different from any "average therapy brand X", with the possible exception of its metaphor having intuitive appeal for a self-helping user. Nor do any of the claimed benefits of a "parts"-focused model not apply equally to approaches based on rules and learning.

I'd like to go a bit meta before I respond to the object-level content in this comment, because I feel like we're talking past each other somehow.

Could you say what exactly is the position that you are arguing for, in this conversation?

For my own behalf, I'm not trying to say that IFS is necessarily the best system, a totally unique system, or even the only system that one should use. I do think that, of the mind hacking systems that I have encountered, it is a pretty good one; and I count it as among one of the most life-changing ones that I have found.

But this is certainly not a claim that it would be impossible to do even better. In fact I'm quite certain that there are some issues for which IFS does not work very well, and for which there must be something better.

So when you say things like, for example,

But none of these arguments actually favor IFS over other systems, or even distinguish it at all! [...] But that "positive intention" frame isn't new, and definitely isn't unique to IFS. Other therapy modalities had it long before IFS was developed, and you can use a reductionist modality without losing the benefits.

then I'm left a little confused as to why you are saying them. I don't think I've said that the "positive intention" frame would be new, or unique to IFS. Certainly there are other therapy modalities that have it too. And these arguments might not favor IFS over other systems, but then I never said that IFS would be the best possible system. I just said that it seems to be a pretty good system that has worked pretty well for me, and for several other people I know.

In terms of conversational frames, I feel like I came to this conversation mostly in a gears-oriented frame, where I don't have any very strong agenda. You seemed to have criticisms about IFS which looked to me like they were based on misunderstandings of IFS, so I tried to correct them; and you seemed confused about why people found IFS valuable, so I tried to share my perspective on why people do. Where we disagree, I'm mostly interested in fleshing out the details of why and how, so as to better combine our models.

And maybe I'm just totally misreading you, but I get the vibe that you are (or at least perceive me to be) in some kind of a dominance frame, and want to establish that IFS isn't a very good system? Or shoot down my claim that IFS is a uniquely good system? Or something?

Could you say what exactly is the position that you are arguing for, in this conversation?

My inner pjeby is arguing for something like this:

"IFS is a specific model that has some similarities and connections to the best available view, but also clear failures that the best available view doesn't have. But giving IFS this much airtime only makes sense to the reader if you think IFS is the best available view! So you should either justify the implicit respect you're giving to IFS, or explicitly acknowledge the source of that respect (like by pointing out that IFS isn't the best available view, but you like it)."

Consider also my comment elsewhere; my sense is that the post is ambiguating between "IFS is of historical interest" and "IFS is how someone should begin to understand this topic" in a way that misses pjeby's meta-level criticisms of the message sent by giving IFS so much attention / not directly agreeing with criticisms. I came away thinking that you think pjeby is right but some part of IFS is worth salvaging, and when I put my pjeby hat on, I can't figure out which part you think is worth salvaging. There are some hypotheses, like that you have gratitude towards IFS, or you think enough rationalists are familiar with IFS that presenting new models like UtEB as diffs from IFS can help them understand both better, or the label for 'theory of modeling psychology' in your mind is 'IFS' instead of something broader, but the actual motivation is unclear from the post.

I came away thinking that you think pjeby is right but some part of IFS is worth salvaging, and when I put my pjeby hat on, I can't figure out which part you think is worth salvaging.

Thank you, yes. That is basically my position as well, though it is also wrapped in a shell of "WTF am I being named in the title and throughout the body of this post when it's actually about Kaj's position and appears only tangentially related to what we talked about before, because although I did answer some questions people asked about what I do, when I was speaking to Kaj I was mainly talking about the distinctions available between IFS and UtEB, not distinctions between IFS and what I, personally, do with clients." So this not only feels like a weird and confusing post on the level you explain, it also feels like an attack on a strawman version of what I do, because Kaj appears in some places to have confused my generic discussion of "deliberate vs. accidental reconsolidation" with me saying something about my own, personal methodology.

Thank you to you and Vaniver for clarifying. I'm sorry for misunderstanding and -representing your position (and for everything else); from now on, I will explicitly interpret all of your UtEB-flavored statements as referencing a generic UtEB-style system, rather than any personal system of yours.

So to answer this. I wouldn't say that any any of the three points in your earlier characterization of my position (meta-schema existing, IFS being able to arrive at the same place as other models, or IFS' frame of positive intention) quite capture the reason why I'm arguing in favor of IFS.

Rather, I do think that IFS provides the best package of applications... of the ones that I personally have encountered so far. I'm not claiming that it would be impossible for a system to be better or that there would be any in principle reason that would make IFS (or parts-based systems in general) intrinsically superior than any others. I'm sure that for every thing that IFS does, there's a way to do it some other way.

Still, of the ones that I have tried, I have found IFS to have the best combination of flexibility, power, and ease-of-learning so far. Maybe I'm just totally ignorant about this, and you tell me of a better one, and then I look into that and start promoting it instead. (I would be very grateful if you did!)

But to compare it with other systems that I've tried / looked into:

  • More mainstream therapies, such as Cognitive Behavioral Therapy and Acceptance and Commitment Therapy. Read a few books on these and definitely found them valuable, but they felt mostly counteractive so limited in their usefulness.
  • Focusing. Useful, but it was often unclear what exactly I should do with the results that came up, and I didn't always get clear answers. Felt relatively laborious.
  • Core Transformation. The first mindhacking technique that I thought *really* worked, and I kept using it for a long time, but some serious issues it seemed to just completely fail with. Also felt like I had to keep repeating uses of it, or the results would fade.
  • Steve Andreas's self-concept editing. Very transformative on a specific set of issues that made me feel like there was something fundamentally wrong with me. However, didn't seem to work on other kinds of issues, such as putting excessive probability on other people thinking bad things of me and deciding to shun me as a result.
  • Internal Double Crux. Found this very useful for dealing with internal conflicts, but it couldn't seem to deal with conflicts involving what IFS would call extreme parts; trying to deal with them produced odd stuff which IDC didn't tell me how to address. This made me look into IFS, which did.
  • Meditation. Seems useful at spontaneously bringing up and healing some issues, but not good for targeted investigation and healing of any particular one.
  • Coherence Therapy. So far I've only read UtEB, which was more focused on giving examples of it than really documenting the system. I've ordered an actual manual, but what I mostly got from UtEB (besides an improved theoretical understanding) was a few extra tools, which generally didn't enable me to do that much new. Coherence Therapy also seems harder to apply, because it requires figuring out what would count as counterevidence for a particular schema, whereas IFS seems to reliably achieve reconsolidation without needing to understand this on an equally explicit level.

So for each of these systems, either they have relatively narrow applicability, or require a lot of experience to use effectively.

In comparison, IFS feel has felt like it has broad applicability and like it is relatively easy to learn. Of course, I had the benefit of having worked with several similar systems before, which no doubt made it faster and quicker for me to figure it out. But a lot of people do seem to take to IFS intuitively. And it feels like IFS is unusually versatile in at least two respects.

First, it's broad in what kinds of situations it can be used for. It gives you a set of skills that lets you

1) Dig into the core of schemas which are based on an incorrect generalization from the original evidence (solving a "problem" which isn't actually a problem in the first place), the way that e.g. Coherence Therapy does

2) Take schemas which are responding to a correct problem with a counterproductive strategy, and update their strategy on the fly (either before going to a situation that would trigger them, or right after they've triggered)

3) Take schemas whose information may or may not be correct, and unblend from them enough that any new information in the situation allows their information to reconsolidate.

4) Mediate internal conflicts arising from schemas which are both correct but contradictory until you reconcile them, the way Internal Double Crux does

Having a collection of four entirely different contexts in which essentially the same skills can be used, seems to make it a lot more flexible than the other, more specialized systems.

Second, my experience is that if you manage to access the original memories behind a schema and do it from a place of Self (two criteria which can admittedly be frequently tricky to get right), then reconsolidation will basically always happen. Like I suggested in the OP, my model is that IFS does this by essentially hacking how extreme reactions are neurally encoded and exploiting the fact that the problem in any schema bottoms out at "and then I would feel so horrible as for it to be totally unbearable", allowing you to reconsolidate that by witnessing it from Self.

This means that IFS works on pretty much any issue where you manage to get that far. IME, something like (say) self-concept editing works on things that make you feel like a horrible person, but not on (say) things where you are afraid of being left alone through no intrinsic fault of your own. But if you have an extreme fear of either one, it's because your brain has a schema which predicts that one of them happening or being the case will cause unbearable suffering. IFS lets you reconsolidate that prediction regardless of what the exact flavor of the problem is. This also seems to be in contrast to say Coherence Therapy, which AFAICT targets the belief one step earlier in the chain. That is, it targets "being a horrible" person or "being left alone", rather than the "and that would cause unbearable suffering" which follows from that. As a result, Coherence Therapy requires figuring out the exact nature of the counterevidence needed, whereas in IFS just witnessing the schema's prediction from Self acts as a universal counterevidence to the prediction of unbearable suffering.

(Probably obvious to you, but just to make it explicit for the other readers; this is not the same as wireheading. You can heal the extreme fear of being left alone, while still strongly preferring not to be left alone and working towards preventing that. What does get fixed is having such a strong fear that you can't reason about it rationally, and extreme reaction patterns which at worst contribute to the very problem they are trying to prevent.)

Of course, none of this means that IFS would be perfect or that I would have managed to fix all of my issues with it alone... but even granting the weaknesses which we've discussed so far, the combination of it being versatile, powerful, and relatively easy to learn makes it the best overall system that I've found. It's also the one which seems to deliver the most "bang for the buck", in case I was forced to choose just one system to teach to other people.

But again, if there's some even better system out there, I'd be happy to be pointed to it! I just haven't found one yet.

Also, I don't know whether this intrinsically requires thinking in terms of parts. Probably you could do all of the same things with a more UtEB-style approach as well. For example, I'm guessing that the reason why the NLP phobia cure procedure works, is that it has that same element of "realizing that you can recall/re-experience this without it being unbearable" as witnessing something from Self does; and there you work on the level of memories rather than parts.

But at least my feeling has been that the "parts interface" gives you the kind of a natural UI from which all of these things flow relatively naturally; if you successfully teach someone the basic set of IFS skills for doing one thing, then it's just a short step towards learning the other things as well. Whereas if you were thinking in more mechanistic terms, you would need more explicit figuring out how to implement each piece. E.g. UtEB only talked about the kind of stuff you do in conventional therapy sessions, and didn't say anything about on-the-fly updating or using the system for decision-making, suggesting that their framework didn't lend itself to those applications being easily invented.

But then again, maybe this is discussed in some CT manual which I haven't read yet. And in honesty's sake, one close friend of mine who has been using IFS for as long as I have, seems to recently have been finding the UtEB approach more effective. So it's certainly possible that I'm wrong about all of this. And again, I would certainly like to be shown a system which was even better than IFS is. :-)

See, now this comment would have made a great article. ;-) I think it says more clearly what you mean than the article you actually wrote, and makes a much better case for your position.

Thanks! Though in all honesty, now that a few days have passed since I wrote the comment... I've been paying more attention to what I actually do currently, and it feels more UtEB-ish in style, so it might not have been correct to say that UtEB didn't enable me to do that much new.

... maybe. It might also be the case that since I never experienced my parts as particularly anthropomorphic, a large part of my IFS has actually always been working directly on the level of memories. And the reason why I thought that UtEB wasn't telling me that many new things in terms of concrete practice, was that my "IFS" had actually been more Coherence Therapy all along.

I'm just confused about what exactly I have been doing, now. :-)

Mostly, I'm confused as to why this is a post at all (especially frontpage); it reads as an attempt to defend IFS against my criticisms of it, without actually disagreeing with or refuting any of those criticisms. In which case, why post it at all?

You seemed to have criticisms about IFS which looked to me like they were based on misunderstandings of IFS, so I tried to correct them; and you seemed confused about why people found IFS valuable, so I tried to share my perspective on why people do.

I'm still trying to find out where you think I have misunderstood IFS. Your explanation of what you found valuable was useful, though, in that I can see the specific insights or experiences you are crediting to IFS that I was blind to because I consider them minimum requirements for a functional system, rather than bonus features of IFS in particular.

But that was largely answered in the comments of the previous thread, so as I said, I'm kind of confused by a frontpage post that seems to be positioning itself as a refutation of my criticisms, but isn't.

I feel like I came to this conversation mostly in a gears-oriented frame, where I don't have any very strong agenda

Me too... to the original conversation. This post seems like an odd escalation on your part, to move from comments to a frontpage post specifically naming me and making it less a quiet conversation between you and I to a public dispute worthy of frontpage attention. It's particularly puzzling because the post spends a huge amount of time vehemently.... not quite disagreeing, but disagreeing enough that a casual observer might think, "Oh, Kaj really gave PJ's ideas a smackdown", and interpret the situation through a dominance frame, even though in actuality you've barely disagreed at all. (In which case again, why post it? What new information does it actually add?)

So a lot of what you're seeing from me is puzzlement. If you were actually trying to inform me of a misunderstanding, this doesn't seem like a good way to go about it. But on the other hand, you haven't actually informed me of any misunderstandings in this post, and instead presented what looks like a weak promotion of IFS and some fully-general counterarguments. In which case, why bring me into it in particular?

I honestly can't make heads or tails of why this is on the front page and not either continued comments on the other thread, or a private conversation, as it doesn't seem to make sense to me no matter how I look at it, based on the intentions you've stated. Instead, it looks like an unnecessary (and possibly hostile) escalation. Rather than impute such intentions to you, my comments have been more trying to figure out what it is you were thinking, using my best guess for the apparent intention of the post (i.e. promote IFS/defend it against my criticism).

The idea that this post was to further mutual understanding is kind of beyond my comprehension, since if that were the case I would've expected you to ask questions in the post rather than simply stating your case. (Or in some parts, just assuming your case rather than even stating it!)

If I recall correctly, I frontpaged the OP.

In case you're not aware of the way that posts get to the frontpage on LW 2.0, when users make a post, it first becomes a personal blogpost, and then the mods make a decision about whether or not to move it to the frontpage. This is a question of whether the post is on-topic for LW (which is defined fairly loosely but mostly just avoids politics and things about in-person rationality communities) and whether it is violating any important internet norms (e.g. doxing). I wrote a post about 2 years ago explaining the basics, though it's fairly long and at some point I should write a shorter version. The way you write about the frontpage sounds to me like you believe the current frontpage/personal distinction mirrors the old Main/Discussion distinction, where users themselves choose to submit things to Main when they think the posts are especially worthwhile to be read. We decided to change that, in large part due to the way posting on Main slide into having higher and higher expectations and caused people to stop writing as much, and even with all the new energy that's risen out of LW 2.0 I expected that eventually the same forces would cause the site to follow that trajectory again.

Specifically for this post, I have not read most of it yet, but before reading your comment I did not at all consider the hypothesis that the post was attempting to escalate a dominance conflict, mostly because I don't think I've ever seen Kaj do something like that in a decade of writing on LessWrong. When I initially saw the OP, I put it in the reference class of many other posts that reply to individuals (e.g. 1, 2, 3, 4) which have generally been received positively or at least neutrally, and frontpaged it.

As an aside, I've been very excited reading your LW comments lately PJEby, I remember reading your comments and posts as they were written back when I was 14-15. I'm sorry you had a jarring experience being named a bunch in the OP.

Edit: Cut some unnecessary stuff i.e. things not about the decision to frontpage this post.

I'm sorry you had a jarring experience being named in the OP.

Thank you. It is at least good to know that it was not his decision to put this on the front page, though the number of times I'm named still makes it feel a bit like it's a calling out, especially since he refers to "pjebyan" practices as if they were what we discussed, rather than the material from UtEB that he himself previously posted.

A lot of what we talked about in the original comments was actually what UtEB describes as reconsolidation, not what I do, because I specifically did not want to get into that here.

Rather, my direct discussion with Kaj was strictly focused on the reductionism issue with parts-oriented models, and the difference between deliberate reconsolidation (ala UtEB) and accidental reconsolidation (ala IFS). It was never supposed to be a referendum on my approach to working with clients or comparing my approach with IFS, outside of me mentioning some reasons why I don't like to use parts-oriented approaches (like IFS or any of its many predecessors), and how my experiences relate to what's said in UtEB.

Indeed, the only reason I felt safe to discuss what I did in that previous thread was because I could use UtEB as an example of a reconsolidation-oriented approach other than mine, because I did not wish to create an impression of using LW as a pulpit from which to preach my own gospel. The unexpected combination of "suddently frontpage" and "naming names/ascribing positions" was quite unpleasant, as it made it feel like I was being shoved into a frame of doing that in direct opposition to my attempts to keep the previous discussion focused on general schools of thought (e.g. behaviorism vs. "parts", deliberate vs. accidental reconsoldiation, etc.) rather than being about "my way is better than yours".

After all, as the guidelines say, "aim to explain, not persuade".

(That being said, I can also see how the frame shift probably seems way more visible and salient to me than it does to anybody else, and on a re-read of the article, even I can see that the parts that got me upset are really very tiny in comparison to the whole. It's also pretty understandable in retrospect why Kaj could easily have thought I was arguing for a model of my own, rather than speaking generically, without him having any intention to distort my views or attribute his own views to me... even as it's also understandable why the situation inclined me to give more weight to the reverse hypothesis.)

Oh. I'm sorry that turning this into a post made you feel uncomfortable and like I was singling you out for public criticism. It wasn't intended as either, but now that you pointed it out, it's kinda obvious that I should have realized that you might and that I should have asked you first.

This post was intended as a compliment in spirit: that I thought your points were important enough that I wanted to highlight them in a more public manner.

I'm not sure if I've ever told you, but I have both high respect for your work in general, and gratitude for your contributions personally. First, you've been talking in a lot of detail about mind hacking theory which has only become obvious to me many years later; probably a lot of what I'm currently slowly puzzling together is stuff that's already obvious to you.

Second, because you mentioned Core Transformation on LW a long time ago, I bought the book at the time; then I didn't get around reading it for a really long time, until I happened to have hit a personal bottom. At that point I finally read it and got a lot of help from it. That then started a chain of events which caused me to find more of the Andreases work, which in turn helped me finally fix the cause of a decades-old depression. None of that might have happened without those old comments of yours, so I'm very grateful for them.

So a part of the motivation for making this post was that I felt that you'd made a lot of good comments and criticisms, and you certainly know what you are talking about, so I wanted to highlight some of those points.

To try to trace back the exact chain of thought that led me to make this post:

  • I started writing a response, and noticed it was getting pretty long, an article's worth on its own
  • I thought something like "pjeby has made several good points on reductionism here; I agree with them, but from the fact that he seems to think I don't, I guess that's not obvious. In case it's just illusion of transparency speaking, I should take these criticisms and more publicly indicate where I agree with them, so that it won't just be buried in the comments of an old post."
  • and "the part about the pragmatic benefits of IFS seems to be getting into a lot of detail about what I think is going on with IFS, as well as new details about how it connects with e.g. the UtEB model; those new details might be interesting to share more widely as well".
  • "and since I just included my summary of what I think to be his core points, and he mostly endorsed them, this would be an easy opportunity to create a distillation of our conversation."

So I made this into its own article (which, as the others noted, I did feel a little uncertain about frontpaging), since that seemed like a nice opportunity to 1) continue our discussion 2) signal-boost and indicate agreement with the points of yours that I thought were important and correct 3) communicate some of my updated thoughts on IFS to a broader audience 4) generally indicate respect for you, in that I'd found your responses important enough to distill and promote.

But I realize now that you found this uncomfortable, and I'm again sorry for not having asked you about it first.

I also didn't realize that you had been trying to explicitly avoid ascribing positions. You had previously discussed your own methods on several occasions, and I assumed that you were contrasting the IFS approach with whatever system you thought was best - and that this would be your own system, since why would you use a system which you didn't think was the best. :-) I'm sorry for misreading and mischaracterizing you; I've now edited the post remove terms such as "pjebyan". If you want me to edit something else in the post, or even take it down entirely, please just let me know.

Thank you for the consideration, and I appreciate the edits. This was just an unfortunate confluence of events and I'm not holding any grudges.

I have to admit that one of my faults is a healthy dose of the illusion of transparency. I tend to assume that other people can reach the same conclusions I have when they have access to the same information my conclusions are based on... even though there's a distinction between say, reading UtEB and grokking what it means about "legacy" approaches to therapy.

So some of the things you said in this article seemed to me like excessively belaboring points I thought were already made quite explicitly in the text of UtEB, so I interpreted it as you trying to argue in favor of IFS, not that you were just now realizing how IFS fit within UtEB's model.

The fact you posted an article about UtEB before made me assume that you understood it at least as well as I did (since in effect, you introduced me to it!), so I didn't see why you would only be now discovering those points... especially since I thought they'd been covered by our previous discussion and your restatement of my position.

Regarding Core Transformation, I'm glad you found it useful. Back at the time I mentioned it, it was one of the better techniques available to me, despite the tendency to sometimes get bogged down in "is that really a part or am I imagining things" or parts getting in circular arguments about things. But I later found that there were simpler ways to address the same things, because what CT calls "core states" are also accessible by simply not activating the parts of the brain that shut off those states. (e.g. by telling us we don't deserve love)

So if, for example, we don't see ourselves as worthless, then experiencing ourselves as "being" or love or okayness is a natural, automatic consequence. Thus I ended up pursing methods that let us switch off the negatives and deal directly with what CT and IFS represent as objecting parts, since these objections are the constraint on us accessing CT's "core states" or IFS's self-leadership and self-compassion.

In effect, you could think of the approaches I've been pursuing since then as shortcutting the process of CT by jumping as directly as possible to our objections to experiencing ourselves as lovable, okay, etc., and working backwards from there.

To put it in context of your changes using Transforming The Self, the shadow qualities (or "negative qualities" as TTS calls them), are the things I target first, since around 2012 or so.

That's because practical experience had shown by then that almost anything I tried to change in myself or others using other methods would often return in a few weeks, unless said negative qualities were somehow addressed. So, strategically, going hunting for them first makes things a lot more efficient, as you then don't have to worry about all the tactical-level behaviors and beliefs being regenerated from the persistent, strategic-level, negative self-image.

Interestingly, now that you've mentioned TTS (indirectly, by linking to your posts referencing it), it reminds me that TTS actually includes something rather like a reconsolidation-oriented approach to quality changes. It might be interesting now to go back and re-read it with our newer knowledge of reconsolidation in mind, to see if I can either improve on his technique, or use something from it to improve on mine.

Thank you, I'm happy to hear that there are no more bad feelings. :)

So if, for example, we don't see ourselves as worthless, then experiencing ourselves as "being" or love or okayness is a natural, automatic consequence.

Cool, I've been having basically the same model for a while. (Related: my hypothesis is that all the talk about people having difficulty "finding meaning" these days seems somewhat misplaced; if things seem meaningless, it's because someone is suffering from objections to their sense of meaning. If those objections would be dealt with, then they would pretty quickly naturally gravitate towards things that felt naturally meaningful.)

Interestingly, now that you've mentioned TTS (indirectly, by linking to your posts referencing it), it reminds me that TTS actually includes something rather like a reconsolidation-oriented approach to quality changes.

Yeah, I don't remember TTS in detail either, but upon reading UtEB it felt like "oh, TTS was a special case of explicitly targeted reconsolidation".

Quick note: Kaj had expressed some uncertainty about whether it made sense as a frontpage post, and a few members of the LW Team had specifically encouraged him to post it as such. So, first, just noting that it probably makes more sense to direct confusions or frustration at us.

I can't speak for the other team members but some of my own thoughts included:

  • the post/comment was long enough that it felt a bit more natural as a post
  • I currently think one of the main bottlenecks on LessWrong is distillation – much of the time, there's a lot of good back and forth in the comments, but it takes a lot of effort for future-people to learn the most important takeaways. The longer the conversation goes, the harder it can be to compress into something digestible. So I think there's something useful to having distillation posts such as this one by Wei_Dai (mostly re: Paul Christiano), and this one by me (re a conversation with Benquo, Jessicata and Zack Davis).

But, I can definitely see how it would come across as weird and escalatory if you weren't expecting it. (I think it is probably a better cultural norm to touch base with conversation partners and give them a chance to give feedback, esp. when you're summarizing their position).

I really appreciate that this post was on the front page, because I wouldn't have seen it otherwise and it was interesting. From an external viewer perspective on the "status games" aspect of it, I think the front page post didn't seem like a dominance attempt, but read as an attempt at truth seeking. I also don't think that it put your arguments in a negative light. Your comments here, on the other hand, definitely feel to an outside observer to be more status-oriented. My visceral reaction upon reading your comment above this one, for example, was that you were trying to demote IFS because it sounds like you make a living promoting this other non-IFS approach.

That said, I remember reading many of your posts on the old LessWrong and I have occasionally wondered what you had gotten up to, since you had stopped posting.

My visceral reaction upon reading your comment above this one, for example, was that you were trying to demote IFS because it sounds like you make a living promoting this other non-IFS approach.

That framing is actually part of what upset me about this article: it presents some of my arguments in a context that makes them seem as though they were made in support of my own approach vs IFS, rather than comparing and contrasting the material discussed by two of Kaj's own posts.

In one post, he presented reconsolidation-oriented therapy as described in Unlocking the Emotional Brain (UtEB for short), and in the other he discussed IFS. My comments in the previous thread were about how UtEB's arguments regarding reconsolidation showcase why IFS is an "accidental reconsolidation" model, and how a deliberate model is more efficient. (Using occasional examples from my experiences with both types of approach.)

This post seems (to me at least) to frame that prior discussion as if I was instead arguing for my methodology vs. IFS, when I was almost exclusively arguing "deliberate vs. accidental reconsolidation", with UtEB from Kaj's own post as an example of the former variety.

So taken out of context, this post makes it sound as if I were doing just what you say: demoting IFS to promote my own approach. But the original conversation was actually comparing two schools of thought that Kaj had written articles about, and by extension, other schools that divide along the same lines.

(But then, my view might be more than a little biased by the unexpected appearance on the frontpage, while thinking that said appearance was Kaj's choice rather than a moderator's, making me look extra-close for why he made a choice that he didn't actually make.)

Yes, that seems like a reasonable perspective. I can see why that would be annoying.

In any case, what I was trying to say was that even if IFS isn’t great reductionism, it’s still a better model than a naïve conception of the mind as a unified whole.

I think there's an important difference between something like "I got insights out of IFS" and "I think people should be taught IFS," and it's not obvious to me whether you're trying to say "historically rationalists got into IFS basically by chance amplified by word-of-mouth recommendations" or "I think rationalists should start learning about this sort of stuff through IFS" or "I think people who got into this through IFS shouldn't abandon it."

This is clarified some by this:

The intention of my comment was descriptive rather than prescriptive; that historically, the IFS model has been popular because it’s pragmatically useful and because, despite its possible flaws, rationalists haven’t been exposed to any better models.

The sense I have from reading through the post is something like "IFS and other approaches seem to be converging, in a way that means differences between them are more subtle / more tied to the overall narrative rather than the active ingredient." If someone is better off thinking of things as parts that have wants, then they should get the IFS-derived version; if not, then not. But this sort of reminds me of my long-standing attempt to figure out what Friston-style models are doing that PCT-style models aren't (and vice versa), where actually the thing at play seems to be a question of "which underlying theory deserves more status?" than "what different predictions do these models make?", in a way that's quite difficult to resolve.

But this sort of reminds me of my long-standing attempt to figure out what Friston-style models are doing that PCT-style models aren't (and vice versa), where actually the thing at play seems to be a question of "which underlying theory deserves more status?" than "what different predictions do these models make?", in a way that's quite difficult to resolve.

As far as differing predictions go, if IFS' model of how parts need to be treated is correct, then it should not be possible to directly alter someone's rules without any reference to parts. Yet, the fact that non-parts-oriented therapies also work is a trivial refutation of this prediction. Ergo, the idea of parts is an unnecessary addition from a theoretical point of view.

If someone is better off thinking of things as parts that have wants, then they should get the IFS-derived version; if not, then not.

Yeah, this is basically my position. I don't have any particularly strong opinion on when exactly one is better off thinking that, though. Feels like it depends a lot on the details of their personality, as well as how interested they are in the details in the first place - e.g. if someone was only interested in getting a practical system that they could use as soon as possible, I'd probably point them to IFS. If they wanted to have a thorough understanding of what exactly they were doing and the neural basis of it, I'd give them UtEB. Or something.

IFS feels like it's easier to just get started with using, since the process is self-guiding to an extent. So I would probably suggest anyone try out at least a couple of sessions with a facilitator, just in case they could grab low-hanging fruit that way. I do think that if people got into this via IFS and it seems to work, no reason to abandon it, though complementing it with other techniques when IFS seems to fail is probably a good idea.

I've been surprised in both directions with different behaviors. 1. That a behavior was part of a surprisingly agentic, self healing pattern with structure that wasn't immediately obvious and 2. That a behavior was a surprisingly structureless spandrel that dissolved on even a modicum of contact with awareness.

So... is there any specific thing that you think cannot be modeled as elegantly as in IFS, by using simple reinforcement learning of when to do a thing or feel a certain way?

Without such examples, I don't think your article has really done anything to improve the case for IFS's validity as an actually-reductionist model of human behavior. That is, when you say:

the “passive” version sounds to me like it’s just a description of how the “agenty” version is implemented.

you're kind of making my point. If we're doing reductionism, then "how it's implemented" is actually pretty important! That sounds like a feature of the passive model, not a bug.

In other words, on at least the "IFS as a reductionist model" side, it looks like you have used an awful lot of words to basically concede my point. ;-)

In other words, on at least the "IFS as a reductionist model" side, it looks like you have used an awful lot of words to basically concede my point. ;-)

Well... yes? That's what I said in the opening, that I think we mostly agree on the reductionist thing but are just choosing to emphasize things somewhat differently and have mild disagreements on what's a useful framing. :-)

Eh, I think that dissolving agency is actually pretty darn important, in a variety of ways, both practical and theoretical. But I think your attempts to salvage the idea of agency are tied to you seeing it as essential to the "positive intention" frame you see as part of IFS' main appeal or value.

So, I've written a separate comment to address that, and the other practical-side arguments of the article.

Eh, I think that dissolving agency is actually pretty darn important, in a variety of ways, both practical and theoretical.

I certainly agree! Have you looked at some of the later posts that I've been referencing in my comments, say the one on neural Turing machines? (I know you read the UtEB one.) Dissolving agency has been one of my reasons for writing them, too.

I'm still not sure what it would mean for humans to actually have subagents, versus to just behave exactly as if they have subagents. I don't know what empirical finding would distinguish between those two theories.

There are some interesting things that crop up during IFS sessions that I think require explanation.

For example, I find it surprising that you can ask the Part a verbal question, and that part will answer in English, and the answer it gives can often be startling, and true. The whole process feels qualitatively different from just "asking yourself" that same question. It also feels qualitatively different from constructing fictional characters and asking them questions.

I also find that taking an IFS approach, in contrast to a pure Focusing approach, results in much more dramatic and noticeable internal/emotional shifts. The IFS framework is accessing internal levers that Focusing alone isn't.

One thing I wanted to show with my toy model, but didn't really succeed, was that arranging an agent architecture where certain functions belong to the "subagents" rather than the "agent" can be more elegant or parsimonious or strictly simpler. Philosophically, I would have preferred to write the code without using any for loops, because I'm pretty sure human brains never do anything that looks like a for loop. Rather, all of the subagents are running constantly, in parallel, and doing something more like message-passing according to their individual needs. The "agent" doesn't check each subagent, sequentially, for its state; the subagents pro-actively inject their states into the global workspace when a certain threshold is met. This is almost certainly how the brain works, regardless of whether you wish to use the word "subagent" or "neural submodule" or what exactly. In this light, at least algorithmically, it would seem that the submodules do qualify as agents, in most senses of the word.

I'm still not sure what it would mean for humans to actually have subagents, versus to just behave exactly as if they have subagents. I don't know what empirical finding would distinguish between those two theories.

I agree that you shouldn't expect to see any findings that distinguish between those theories. But I thought the main question here was closer to "do humans behave as if they have subagents?", where there's evidence that points in that direction ("I have conflicting desires and moods!") and evidence that points away from that direction ("Anger isn't an agent, if it were you would see Y!").

My examples of subagents appearing to mysteriously answer questions was meant to suggest that there are subtle things that IFS explains/predicts, which aren't automatically explained in other models. Examples of phenomena that contradict IFS model would be even more useful, though I'm failing to think of what those would look like.

Examples of phenomena that contradict IFS model would be even more useful, though I'm failing to think of what those would look like.

The reason it's hard to think of what it would look like is that viewing things through an agentic lens makes you miss those counterexamples by confabulation. For example, you can trivially explain any behavior by postulating a part that wants to do that behavior. In contrast, simpler models have to be grounded in something more concrete (such as what the rule(s) are and how they were learned), which means such models are less flexible... and thus more likely to tell us something useful about the world, due to what they rule out.

My examples of subagents appearing to mysteriously answer questions was meant to suggest that there are subtle things that IFS explains/predicts, which aren't automatically explained in other models

I don't see that, though: if you ask people questions, they are mysteriously able to answer them, and the answers can often be startling, and true. While I can't speak to your subjective experience, I can describe something from mine that sounds similar: if somebody asks me a question under certain circumstances, I find myself listening to an answer coming out of my mouth that I did not know before, and which I do not experience myself as knowing until after I hear myself say it.

This state of affairs does not involve IFS-style parts, so ISTM that IFS does not add anything special here to the idea that questions can trigger the appearance of information in the mind that one did not explicitly know beforehand... Almost as if it had just been prepared fresh by a chef in the kitchen, vs. something that was already in our refrigerator of knowledge. ;-)

Disclaimer: I only know about IFS from this sequence, so I might confuse parts of it with my own models.

I think there is a value (or at least could be) in speaking of IFS parts as being person-like to a somewhat larger degree than a fully reductionist model would imply them to be:

When focusing on a part, a big chunk of one's mind is involved, which I expect to lend one's experience of the part actual person-like properties even if they were not there initially. Also, I would expect this effect to be easily amplified by having expectations of agency while focusing on the part (if I expect the part to be person-like, I will model it as a person and then just use my person-model of the part in order to interact with it). It seems plausible/possible that using one's 'dealing with a person'-abilities (where everyone has a lot of experience) for interacting with parts is easier to apply than using a more abstract model (I do not know much about other models, so it might well be that there are other easy-for-humans-to-apply methods which don't use 'mini-people').

With this I think that the IFS framing of parts as 'mini-people' can be seen as a feature, not a bug - although one should keep in mind that perceiving a part as strongly person-like is not necessarily a property of the part. I would expect: For improving one's understanding of the mind, one should not overestimate the degree of person-hood of parts; for dealing with one's parts, treating them as mini-people might be useful even if they aren't.

My impression was that parts in IFS range from if-then-rules to dissociated personalities (I was surprised to learn that their existence is debated; non-debated complex examples are described in Subagents, trauma and rationality). Because of this I thought that the descriptions as 'mini-people' are mostly meant to be easy to grasp and remember and do not claim to be accurate over the whole range.